Do pro roasters cup each batch before blending?

Discuss roast levels and profiles for espresso, equipment for roasting coffee.

#1: Post by OkcEspresso »

I have been curious about this for some time. When roasters are producing blends, do they/you roast each batch, cup immediately, then decide which beans and it what quantities to mix to produce your blend? Do you do this for each batch or do you assume a bag/lot will be mostly consistent and you can just repeat the same mix ratio and produce the same results?

In slightly more detail, for espresso blends, do you perform a traditional cupping post roasting each SO or do you actually make espresso for cupping after some degassing period then use whatever ratios worked? I have assumed that someone with extraordinary taste perception could likely roast, cup immediately and determine most of the flavor characteristics that would be present in an espresso post-degassing. Lately some of my roasts have surprised me in the way that they change over time and my curiosity induced this post.

Since at home, I have no interest in producing the exact some flavor with each blend, I am not actually shooting for anything specific. Some of the roasted beans I purchase vary dramatically from one bag to the next. Some of this could be my fluid barista skills, but I think it is the blending.


User avatar
Team HB

#2: Post by another_jim »

I've seen pro roasters develop blends. It's an ongoing process:

1. Traditional cupping or SO shots in the lab are used to come up with a basic blend.
2. One roasts a lot of variations, trying pre and post blends, tries out the variations over a course of weeks or months, preferably in a store setting and decides on the best one.
3. This process is repeated when new coffee shipments and crops arrive, although here the target is to maintain a successful signature taste.

I do a lot of blending in a home setting, here the process is different. I roast mostly SOs and mix and match on whim. Sometimes, usually once a year or maybe once every two years, I come up with a good blend that can be premixed. I'll then mix up 3 to 5 pounds of it, send out samples to friends, get feedback, and finally make up a tweaked version of as much as my stash will bear. When I don't have a complete blend, I'll mostly do a half-blend: that is a mix of basic brazils, dp Ethiopians, and indos. To this I'll add 10% to 20% of whatever very light roasted coffee I'm currently brewing as an aromatic top end kicker to give it some interest.
Jim Schulman

BPlus: turning your coffee spirit
Sponsored by BPlus
OkcEspresso (original poster)

#3: Post by OkcEspresso (original poster) »

Thanks Jim for that insight.

When the pros are then mixing up production batches are they cupping each roast batch to make sure it is within the taste specs they are shooting for? Or do you think they just roast per the established paramaters they used to design the blend and blend per the same parameters? In other words, are they tasting every component of every batch?

If so, are they making adjustments based on the cupping results to attempt to get the "correct" final blend?

User avatar
Team HB

#4: Post by another_jim » replying to OkcEspresso »

The roasters whose practice I know cup every roast, single origin or blend, to make sure everything is going right. For regular coffee, where there's no resting period before shipping or stocking, even cupping the same day can be a bit late. George Howell at Terroir once "recalled" the early sales of a roast where the blower on the roaster malfunctioned; I thought that was very cool. I'm sure other roasters have done the same thing in response to glitches.
Jim Schulman


#5: Post by PeterG »

Very interesting question.

Roasters have various tools at hand for quality assessment/control, only one of which is cupping.

But, as for your questions:

Green coffee: We pay a lot of attention when buying coffee to potential variation within a given lot; i.e. bag-to-bag variation. This is one reason good coffee buyers travel to inspect mills, and insure that coffees are sufficiently consistent at the milling/bagging stage. We make every possible effort to eliminate bag-to-bag variation. When coffee is from a good mill, you can count on the coffee being consistent throughout the lot.

Blend creation: usually, by the time we are actually roasting coffees, we are intimately familiar with a given coffee, and are starting to predict how it will behave in darker roasts, the espresso machine, etc. After a while, you begin to be able to predict how coffees will interact in a blend. Also, frequently, you wind up buying coffees with blends in mind (this happens often for espresso). In any case, just as Jim said, we usually come up with espresso blends based on basic information (cupping and single origin shots) followed by a number of experimental blends. Once a blend is hit upon, you tweak it frequently over the course of the year to maintain the integrity of the flavor profile. This is not done every roast, more like every few weeks or even more.

We maintain consistency by using a spectrophotometer to analyze roast color of both beans and ground coffee after each roast. Weight loss, roasting time and temperature can also give you valuable information. Using all of these things, we can accurately predict the flavor of the coffee and therefore how it will behave in a blend.

You actually need to be somewhat careful about how much you cup. Even the most accurate palate varies from day to day and over the course of the day, so if a roaster were to make adjustments based on individual cuppings of individual roasts you would be all over the place. In addition, cupping "right out of the roaster" is almost useless, and good roasters generally ship their coffee out before it is sufficiently degassed to give good cupping results. For this reason, data collection, followed by corresponding those data with cupping results is most valuable.

Peter G
counter culture coffee

User avatar
Team HB

#6: Post by another_jim »

Hi Peter,

You're a lot more scientific than the roasters I've seen (unless they are hiding their cool toys :wink: )
Jim Schulman

OkcEspresso (original poster)

#7: Post by OkcEspresso (original poster) »

Thank you Peter for the detailed explanation. I wonder if a spectrophotometer is the norm or only used by exceptional roasters. The level of variation I taste from one bag to the next of some roasters is almost disconcerting. I remember seeing a discussion about this somewhere, but do you guys cull post-roast and what do you cull if you do?